Developmental process of vertebrate embryos is regulated, at least in part, by secreting molecules such as growth factors. We are focusing on the function of Bone Morphogenetic Proteins (BMPs) that are the members of TGF-beta superfamily during mouse development. To reveal the function of BMP signaling, we have generated a mutant mouse that is deficient for BMP type IA receptor (Bmpr1a or Alk3) and activin type IA receptor (Alk2) by conventional gene targeting technologies. Nullizygosity of each receptor caused severe embryonic lethality and mutant embryos die at embryonic day 7.5 (Bmpr1a) or 8.5 (Alk2). For Bmpr1a mutant embryos, we found that 1) no mesoderm was formed, and 2) cell cycles prior to gastrulation was prolonged. For the Alk2 mutant embryos, we found that 1) mesoderm was formed but not fully differentiated, 2) Alk2 signaling in the extraembryonic region (future placenta) was critical for gastrulation, and 3) Alk2 mutant cells were not capable to contribute heart or eye. These results suggest that BMP signaling at the early stage of embryogenesis is important for cell growth, gastrulation and formation of particular organs such as heart.[unreadable] [unreadable] For functional analysis of these gene products in a later stage of development, we introduced a tissue-specific gene targeting. Using this technology, we mutated Bmpr1a in a bone-specific manner (specific for differentiated osteoblasts). The bone-specific Bmpr1a deficient mice were viable indicating we can avoid embryonic lethality of Bmpr1a disruption with the tissue-specific gene targeting technology. Mutant mice were smaller than normal littermate and show less deposition of bone matrix in their bones. These results are the first evidences that BMP signaling is required for normal bone formation in vivo. Interestingly, the aged mutant mice showed reduced osteoclast activity that led to increased bone matrix. This suggests that primary function of BMP signaling in the differentiated osteoblasts is age dependent. To directly address the age-dependent alteration of BMP function in osteoblasts, we employed a tamoxifen inducible Cre system. A Cre recombinase of which activity can be induced by tamoxifen treatment was specifically expressed in osteoblasts using type I collagen promoter (Col1-CreER). When tamoxifen was administrated to newborn pups, bones in the mutant mice showed reduced bone mineral density (BMD) compared to the control littermate. When tamoxifen was administrated to the weaned mice between 1.5 month and 5 months after birth, bones in the mutant mice showed higher BMD than controls. These results are in good agreement with our previous finding and strongly suggest an age-dependent function of BMP signaling in osteoblasts. In another approach, we mutated Bmpr1a in a neural crest specific manner. Embryos die at the mid-gestation stage showing abnormalities in neural crest cell-derived tissues such as dorsal root ganglion and craniofacial region. These embryos also showed abnormalities in outflow tracts (OFT) where cardiac neural crest cells migrated. Preliminary results revealed a reverse flow in the dorsal aorta in E11.5 mutants, suggesting functional abnormalities in the cardiac cushion in OFT. This is the first evidence that the OFT cushion is functional as early as E11.5, and BMP signaling is one of the critical players for cardiac neural crest cells to form functional OFT cushion. We also mutated Bmpr1a in an epiblast-specific manner at the stage of gastrulation. Embryos showed an overgrowth of neural tissues after gastrulation. Paraxial mesoderm such as somite is expanded in the mutant embryos, but no sign of the heart development. These results indicate that BMP signaling plays a critical role in various stages of neural tissue development as well as mesodermal patterning.[unreadable] [unreadable] In order to elucidate the function of ALK2 signaling during the later stages of development, we set up chimeric analyses to determine if ALK2 signaling is essential for heart morphogenesis, eye development, and establishment of left-right identity along the body axis. To set up chimeric analysis for Alk2, we isolated homozygous mutant ES cells for Alk2 (-/-) and injected them into the cavities of wild type blastocysts. The ES cells are marked by beta-galactosidase, therefore, mutant cells can easily be distinguished from wild type cells in the chimeric embryos. In the chimeric embryos, abnormalities caused by Alk2 mutation are partially rescued and they can survive several more days in utero than a homozygous null mutant. Chimeric embryos recovered at E9.0-E10.0 showed a variety of phenotypes depending on the degree of contribution of Alk2 from the mutant ES cells. Chimeric embryos with low contribution of the mutant ES cells (less than 80%) developed without showing overt phenotype. Chimeric embryos with high contribution showed growth retardation and die around E11.0. In normal embryos, a heart tube exhibits rightward looping at E9.0, then forms a four-chamber structure later. Chimeric embryos showed defects in turning and heart morphogenesis; looping of the heart tube was randomized. Nodal, Pitx2, and lefty2 are expressed in lateral plate mesoderm, but their expression is limited to the left side in normal embryos. Interestingly, these genes were expressed on both sides of the lateral mesoderm in the highly contributed chimeric embryos. These results indicate that the chimeric embryos fail to establish left-right asymmetry at the node leading to left isomerism. To get further insights, we have set up an epiblast-specific disruption of Alk2 using floxed Alk2 mice and Mox2-Cre mice. Preliminary results revealed that cilia formation in the node was compromised leading to the loss of the leftward nodal flow.